st: Different results from interaction term and Suest Command

[natasha agarwal <agarwana2@googlemail.com>]

Dear everyone,

I am trying to test the hypothesis whether development level on a
region influences the impact of FDI on the value added of domestic
firms within a region.

For similar reasons, I estimate an augmented production function and
adopt two different strategies to test whether the slopes on the FDI
variable in the two groups are similar.

First Method with the interaction term:

xtreg lnrval lnk lnw lnvfdi fdev1 y14-y18, fe vce(robust)

where fdev1=lnvfdi*dev1 where dev1=1 if region higher developed and
zero otherwise. In this method fdev1 (which is the difference between
the two groups is not statistically significant)

Second Method with sample splitting

bysort number: center lnrval lnk lnw lnvfdi y14-y18, casewise
regress c_lnrval c_lnk c_lnw c_lnvfdi c_y14-c_y18 if fdev1==1
est sto fe2
reg c_lnrval c_lnk c_lnw c_lnvfdi c_y14-c_y18 if fdev1==0
est sto fe3
suest fe2 fe3, vce(cluster number)
est sto fe4
test [fe2_mean]c_lnvfdi==[fe3_mean]c_lnvfdi (the slopes are
significantly different)

The difference in the slopes from the two methodologies is different.

Can anyone please help me with this?

Thanks
Natasha
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/